6i ] the sensory physiology of the harvest ...sensory physiology of the harvest mite trombicula...

[46i ]

THE SENSORY PHYSIOLOGY OF THE HARVEST MITETROMBICULA AUTUMNALIS SHAW

BY B. M. JONES

Department of Zoology, University of Edinburgh

(Received 18 May 1950)

(With Twenty-four Text-figures)

INTRODUCTION

The ectoparasitic habit of the hexapod larva of Trombicula autumnalis is the cause ofmuch discomfort to residents of infected localities in the British Isles, between lateJune and the beginning of October. The mite is a member of the Trombiculid groupwhich includes species known to transmit disease in some parts of the world.

The unfed larvae are found either upon the soil or climbing upon low-lyingvegetation. Under suitable conditions they aggregate into clusters and are thenmore easily detected as orange patches. Development to the nymphal stage cannottake place unless the larvae obtain a meal from the superficial tissue of a vertebratehost to which they must securely attach themselves. The nymphs and adults arenon-parasitic and lead a hypogeal existence at a depth of about 12 in. below thesurface of the soil (Cockings, 1948).

The hairs of a mammal, or the feathers of a bird, as they brush against infectedsoil or low-lying vegetation, are admirably suited for picking up the mites, but thequestion arises, to what extent are sensory perceptions of environmental stimuli ofthe mites directed towards the acquisition of a host. The chief aim of the presentwork has therefore been to investigate (a) the responses of the mite to stimuli mostlikely to have value with respect to the problem of acquiring a host, and (b) thenature of the sensory organs.

Few workers have studied the orientation mechanisms of members of the Acarina.Henschel (1929) described the reactions of Tyrolichus casei, the cheese mite, tochemical stimulation. Totze (1933) studied the sheep tick, Ixodes ricinus, and Solo-mon (1937) the red-legged earth mite, Halotydeus destructor, in relation to environ-mental conditions. Lees (1948) reinvestigated the reactions of Ixodes ricinus withrespect to those stimuli the tick will encounter in its natural environment. As to thebehaviour of harvest mites there exist only a few scattered and very incompletereferences to observations in the field. It was found desirable to carry out a pre-liminary ecological study of Trombicula autumnalis in 1947 to provide informationupon which to base the present work. The observations of the mite in the field wereinvaluable for suggesting lines of investigation in the laboratory, and for assessingthe significance of the sensory perceptions of the mite with respect to acquiringa host.

JEB.27, 3*4 3°

462 B. M. JONES

REACTIONS TO LIGHT

Among members of the Acarina, unfed ticks and harvest mites closely resembleeach other in showing a tendency to climb upwards, a movement which in someunfed insect larvae is associated with a positive response to light. However, theresults of Totze (1933) and MacLeod (1935) on the response to light of eyelessspecies of ticks are contradictory. Lees (1948) found himself in agreement withMacLeod, who was quite unable to confirm the findings of Totze, who maintainedthat unfed ticks, Ixodes ricinus, in all stages are strongly photopositive. Lees, likeMacLeod, tested unfed and engorged ticks which were all photonegative.

Ticks show a strong inclination to climb up to the tips of the stems of rushes andgrasses. This behaviour could be accounted for either by a response to light, or to theinfluence of some form of negative geotaxis. Krijgsman (1937), working with ticklarvae, Boophilus annulatus, found they were indifferent to gravity. Lees (1948)stated that it was not easy to interpret the results of his gravity tests on Ixodes ricinus,but concluded that although negative geotaxis may be of some significance, theinclination to maintain a position at the tips of glass rods, serving as models ofnatural grass or rush stems, was partly a tactile response following arrival at the tip.If, as the evidence suggests, ticks are photonegative and independent of gravity, it isdifficult to account for the persistent nature of their tendency to climb, upwards,which must invariably lead unfed ticks, in the first place, from the roots of vegetationand up the stems of grasses, before they experience arrival at the tip, which as Leessuggests only enhances this behaviour.

The findings of gravity tests on harvest mites (see p. 482) showed that they assumeda random distribution upon a vertical rod. This apparent independence of gravity isalso readily shown by the mites when they move about upside down upon the undersurface of the window of a light trap (Jones, 1950 a). With the elimination of gravityas an influence, there remained the possibility of a response to light being responsiblefor the upward-climbing movement.

Responses to beams of light

Method. The tests were made in a dark room. A circle of filter-paper, 18 cm. indiameter, placed upon a circular glass plate, 20 cm. in diameter, served as theexperimental field. Although a dark surface is usually prescribed, preliminary testsshowed the reactions of mites to horizontal light to be similar upon either black orwhite filter-paper. White filter-papers were therefore used because it was easier tomark the paths traversed by the mites. Illumination was provided by 40 W. and100 W. bulbs enclosed in light-proof containers with a circular aperture "z\ cm. indiameter, to allow the escape of a horizontal beam of light across the filter-paper.The light, before it escaped, was first cooled by passing it through a z\ in. tankcontaining an acidified solution of alum. Light intensities were measured witha G.E.C. photometer.

Single horizontal beam. Mites were placed, usually three at a time, in the centre ofthe illuminated field. On exposure to a strong beam, with a light gradient extending

Sensory physiology of the harvest mite Trombicula autumnalis 463

from 670 m.c. at the centre of the field to 6500 m.c. at the source, they usuallymoved away from the light before tracing a circular path either to the right or left,which brought them into a position facing up the gradient. This initial turningmovement was a prelude to the mites tracing paths straight towards the source, butwith a tendency to curve to one side of the beam (see Fig. 1B). The movements ofmites exposed to a weaker beam, with a light gradient extending from 270 to2600 m.c., did not show the same features. The initial turning, induced by thestrong beam, was not evident, and the tracks towards the source were decidedly wavy,and only straightened near the source (see Fig. 1 A). Occasionally a mite movedaway from the light and when transferred back to the centre of the field it repeated

-670 m.c.

Fig. 1. Tracks of harvest mites in a horizontal beam of light. A, in a beam gradedfrom 2600 to 270 m.c.; B, in a beam graded from 6500 to 670 m.c.

the negative response. Sometimes a photopositive mite when repeatedly broughtback to the centre of the field orientated towards the light by taking almost the samepath it had traced previously. Behaviour of this kind has probably been responsiblefor the suggestion that each animal inherits a specific response to an offeredstimulus.

Two horizontal beams. In a two-light experiment with beams of equal intensityarranged to intersect at about 900 at the centre of the field, the mites were inclined tofollow a path some distance along the bisector, before moving towards one or theother light source (see Fig. 2B). When two unequal crossed beams were presentedthe mites moved directly up the gradient of the stronger beam (see Fig. 2 A).

To discover the ability of the mites to detect a stronger light intensity behind them,they were allowed to move up the gradient of a single beam until they had approachedwithin about 6 cm. of the source (2600 m.c), before an opposing light of 6500 m.c.at the source was switched on. The mites continued in the same direction for abouta centimetre, stopped, made a complete turn, and then moved towards the stronger

30-2

464 B. M. JONES

light. Occasionally a mite displayed continuous turning movements on reaching thecentre of the field (see Fig. 3).

Fig. 2. Paths followed by harvest mites in two horizontal beams of light which intersect at rightangles. A, in two unequal lights, one of 6500 m.c. CZt)> the other of 2600 m.c. (-»-), B, in twoequal lights of 2600 m.c. (-*•).

Fig. 3. Tracks of harvest mites between two unequal lights. The opposing light of 6500 m.c. ( *)was switched on when the mites had moved within 2 or 3 cm. of a light of 2600 m.c. (->).

Response to a laterally presented beam

In this series of experiments the light source remained in the same position,whereas the filter-paper was turned about a centre point like a record on a gramo-phone disk. The mite was placed in the beam and allowed to move towards a light.The filter-paper and the mite were then turned clockwise through an angle of 90°.The light now fell laterally upon the mite so that one eyespot was more stimulatedthan the other. This asymmetry of stimulation caused the mite to re-orientate itselftowards the light.

By repeatedly turning the filter-paper in this way the mite was induced to tracea rectangular path which brought it back to the starting point (see Fig. 4). However,individual mites showed variations in. their reactions. The turning began eitherimmediately or after a slight delay. Occasionally a mite after each turning movementtraced a path directly towards the source of light. More often they moved towards it


at an angle of about 150, but they corrected the deviation if they were allowed tocontinue to approach the light source.

12

Fig. 4. Response to lateral illumination. The beginning of each track of the harvest mite is denotedby the circle, and the position of the light is indicated by the arrow numbered i ; thereafter thelight was presented laterally.

Blinding experiments

Mites previously chilled to make them inactive were transferred to a moist filter-paper under a binocular microscope. Quick-drying black cellulose paint was placed,with the end of a single camel hair glued to a wooden handle, over the ocular area.Treatment was considered successful if the paint completely covered the areawithout interfering with any part of the body concerned in locomotion. The treatedmite was then transferred to a glass cell and kept in complete darkness for an hour.

When mites blinded on one side were placed in a uniform light of 250 m.c.provided by an overhead lamp, they moved in the direction of the seeing eye, andmade circus movements as they traced their somewhat random tracks towards theperiphery of the experimental field (see Fig. 5 B). Unblinded mites made more orless straight tracks from the centre to the periphery of the field.

270 m.c.

2600 m.c.

Scm.

Fig. 5. Tracks of harvest mites with the left eye blinded. A, in a horizontal beam of light;B, when exposed to an overhead light.

When unilaterally blinded mites were exposed to a horizontal beam, they wereinclined to make a gently curving circus movement before tracing a path diagonallyacross the beam and into the shaded part of the field. It was clear that the primaryeffect of the light was to stimulate the seeing eye, and increase the muscular tone ofthe uncovered side sufficiently to cause a deviation to one side as the mite approachedthe source (see Fig. 5 A).

466 B. M. JONES

The response of. totally blinded mites was also tested to discover whether or nota diffuse dermal light sense existed. In the way already described, mites with botheyes blinded were exposed to a horizontal beam of light, but, it being exceedinglydifficult to accomplish the satisfactory blinding of both eyes of an individual mite,the number of tests was limited. The results of these tests (see Fig. 6) are noteasy to interpret, but it is clear that the blinded mite completely loses the ability tomove directly towards the light source. Apparent movement towards the light,when it occurred, could be accounted for by the possibility of light entering the thincuticle, and reaching the eyes from below (Oehring, 1934, after Wigglesworth, 1939).This is certainly a possibility because mites, contained in a glass jar exposed to light,will take up a position nearest to the source, although this entails the ventral side oftheir bodies facing the light (see p. 481). The same response is equally well demon-strated by their response to the horizontal window of a light-trap, where the mites,walking upside down upon the under-surface of the window, remain in the incidentlight (Jones, 1950 a).

Fig. 6. Tracks of harvest mites with both eyes blinded, in a horizontal beam of lightfrom a source of 2600 m.c. (-»•).

Reactions to light intensity

The experiments were designed to discover whether the mites showed a selectiveactivity towards light intensities under conditions in which the source of light wasnot apparent.

A glass tube, 27 cm. in length and o-6 cm. in diameter, closed at each end bya ground-glass stopper, was marked off into three equal lengths A, B and C.Section A was blackened with a deposit of Sudan black. The middle section B wasshaded by an opaque screen to give a light intensity of about 300 m.c, whilst thelight intensity of section C, exposed to the northern light of the window, was about500 m.c. Fifty mites were placed in the middle of the tube for each test. The mitesscattered along the tube, and whereas they avoided the darkened section there wasno indication that they showed a preference for the window light. The mites were


continually active and moved to and fro indiscriminately between sections B and C.The movements of individual mites are shown in Fig. 7.

To accentuate the difference between the light intensities the glass tube was placedin direct sunlight. A moist pad of cotton-wool was inserted into the tube near theclosed end of the darkened section to provide a favourable air humidity inside thetube. The mites still avoided the darkened section, but occasionally they wanderedinto it for a brief period. Rough positional counts, made every 15 min., of mites inthe shaded section B and the sunlit section C showed that about two-thirds of thenumber of mites tended to remain in the sunlit portion. Tests carried out withindividual mites showed that frequently the mite, when placed between sections B

Fig. 7. Movements of harvest mites inside a horizontal glass tube exposed to different light intensities.A, darkened; B, shaded (300 m.c); C, exposed to a light from a window facing north (500 m.c).

Fig. 8. Movements of harvest mites inside a horizontal glass tube of which section C is exposed tobright sunlight and section B is shaded with an opaque glass screen.

and C, first crawled away from the more intense light. Occasionally it entered thedarkened section for a brief period before making very devious tracks towards thesunlit portion of the tube. But on having entered the sunlit portion the mite remainedthere actively crawling around the inside of the tube (see Fig. 8).

Such experiments, in which a portion of the tube containing mites was exposed todirect sunlight, did not allow for the possible influence of temperature and humidityupon the observed distribution of the mites. Extremes of light intensity were there-fore offered to mites in a tube in which temperature and humidity variations wereeliminated. The tube was marked off this time into four sections, A, B, C and D-Section A was darkened; B (300 m.c.) was shaded by an opaque screen; C (500 m.c.)was open to the window light; and section D (1000 m.c.) was exposed to a water-

468 B. M. JONES

cooled beam of light (see Fig. io). The beam was screened from the other sections ofthe tube. The air temperature in each marked off portion of the tube was 190 C.

5 cm.

Fig. 9. Tracks of harvest mites inside a horizontal glass tube divided into four sections of differentlight intensities. A, darkened; B, shaded (300 m.c.); C, exposed to a window facing north(500 m.c); D, exposed to artificial light of 1000 m.c.

A, 5 cm. ,

l . J C E

1

1 I 1

i B ! C v

Fig. 10.

1 5 cm. ,

Fig. 11.

1 •'• 1 M

/y\1,3 cm.

1 1 1 1

0

D

r ;

1Fig. 12. Fig. 13.

Figs. 10-1.3. For explanation see text.

Tests were carried out with individual mites. In most of them the mite movedstraight towards section D and remained there. When the ground-glass stopper wasremoved at this end of the tube the mite displayed a well-defined preference for


crawling actively around the rim of the open end. The mite was usually removedafter it had crawled on the rim for 20 min. or more (see Fig. 9). This response to thehighest intensity of light indicated that the response to sunlight was a true responseto light and not to heat. The avoiding reaction to darkness was the natural outcomeof the photopositive mite orientating itself towards the lighter part of the field.

When mites are transferred to a collecting jar exposed (either in the field or in thelaboratory) to daylight they become aroused and display an immediate upward-climbing reaction. They aggregate round the base of the glass stopper and penetrateany space between the stopper and the neck of the jar. If the stopper is removedthey will crawl round the rim of the neck, but they are just as liable to climb directlyover the rim, down the outside of the jar and to the substratum. The upward-climbing movement was not the result of an avoiding reaction to condensed moisturefrom moist filter-paper at the bottom of the jar, because the neck was kept moist witha wet camel-hair brush each time a fresh batch of mites was introduced. Thebehaviour suggested either a response by the photopositive mite to sunlight ornegative geotaxis.

Response to shading

In the field it was not difficult to obtain from the mites a very definite questingresponse to a shadow suddenly passed over them. It was best elicited from a quiescentcluster of mites exposed to sunlight. When the hand was passed rapidly over a clusterof mites they displayed immediate and very definite reflex actions. The cluster wastransformed into a mass of waving legs. The typical reaction of an individual mitewas to direct the anterior part of the body upward and hold the forelegs out in front

0*2 mm.

Fig. 14. Questing response (shadow reflex) of harvest mites. A, the waving forelegs;B, the body lowered and the legs directed upward.

of the body. The forelegs beat the air with alternate up and down movements as ifthe mite was trying to reach something immediately above. The frantic beating ofthe forelegs subsequently stopped and the cluster broke up, with the mites movingaway from the site. A shadow passed over the crawling mites a sceond time stoppedthem in their tracks and induced the typical questing response. When a cluster ofmites was dispersed by passing a shadow only once or twice it was common for themites to reform into a cluster provided there was no further interference. If the

47° B. M. JONES

shadow stimulus was produced repeatedly the questing response gradually weakenedas the aroused mites wandered in all directions away from the original site.

This response to a sudden decrease in the light intensity was observed in moredetail under the microscope in the laboratory. When a shadow, thrown by a needle,was passed over them they immediately stopped in their tracks and the forelegs beatthe air, whilst the body of the mite was supported by the mid- and hind-legs(see Fig. 14A). Frequently the body was lowered suddenly to the substratum,whilst all the legs were extended upward (see Fig. 14B). In the same way the shadowof a passing host would equally induce the mites to quest, a condition of prepared-ness, which would greatly improve their chances of being picked up.

Response to reflected light

Method. The experimental field consisted of an arena 14 cm. in diameter witha wall 6 cm. high. The arena was constructed by placing a collar of black paper upona sheet of black paper which was the floor of the experimental field. A piece of whitepaper, 6 cm. square, served as a white screen against the black wall of the arena.An overhead lamp illuminated the arena with a weak but uniform light of 50 m.c.

Results. At first the mites usually turned away from the white screen before theymoved towards it, but in all cases the screen was eventually reached. The tracks were

Fig. 15. Tracts of harvest mites towards a white screen presented in a blackarena exposed to an overhead light.

very wavy, and those which were inclined to lead towards the side of the screencurved towards it as the distance shortened (see Fig. 15). None of the paths takensuggested that the mites reached the screen by a series of avoiding reactions to thedark surface of the arena, so that this response to a white surface suggests that it isone towards reflected light, an orientation described in many insects (Fraenkel &Gunn, 1940).

REACTIONS TO TEMPERATURE

Temperature is an important factor to insect parasites of warm-blooded animals,and is one of the stimuli guiding the insect to its host. Wigglesworth & Gillett (1934)found that Rhodnius prolixus, the large South American blood-sucking bug, moved


towards a test-tube of warm water, and that the receptors for detecting warmth werelocated in the antennae. Rivnay (1932) showed that the bed-bug is thermotactic.Mosquitoes and some lice are also stimulated to bite by the heat of the body.Totze (1933) mentioned, and Lees (1948) later confirmed, that the tick, Ixodesricinus, will move towards a warm tube. The common occurrence of harvest miteson warm-blooded hosts suggested the possibility of temperature being a guidingstimulus.

Response to a source of heat

Method. The purpose of the experiments was to learn whether a concentrictemperature gradient set up around a heated glass tube could guide the mite towardsthe tube itself.

A flat-bottomed glass tube, 2 cm. in diameter, was connected by a capillary tubearrangement to an adjacent large vessel containing warmed water. The warm watercirculated through the tube at a controlled temperature, and ran off through anoutlet capillary tube. A thermometer was also inserted into the tube to take tem-perature readings. The tube itself was placed in the centre of a filter paper, the mites(previously kept at a temperature of 180 C.) were released at varying distances fromit. The temperature of the laboratory whilst the tests were being carried out was18-190 C.

Fig. 16. Responses of harvest mites to a warmed glass tube. A, tracks around a glass tube warmed to300 C.; B, tracks around a glass tube warmed to 400 C , with a circular track representing move-ment around the base.

Results. Mites released at distances of 1 and 0-5 cm. from warm tubes withtemperatures of 25, 30, 35, 40 and 450 C. showed no response. Occasionally indivi-dual mites in their wanderings from the point of release moved towards the tube andtouched it. At temperatures of 25-30° C. the mites which had moved against theheated tube showed no positive response and continued their wanderings away fromthe tube towards the edge of the filter-paper (see Fig. 16 A). At temperatures of35-40° C. mites which had accidentally touched the tube displayed a well-definedresponse. For a period varying around 3 min. they crawled continually around thebase of the heated tube before climbing upon it (see Fig. 16B). Mites placed againstthe base of the tube at temperatures between 35 and 40° C. invariably displayed thisdefinite reaction of excitedly crawling around the base of the tube. They were not

472 B. M. JONES

repelled by a temperature of 45 ° C , but they crawled around the base of the tubeuntil the temperature had fallen to about 400 C. before climbing upon it.

Since the temperature from which the mites were taken was about 190 C. and thetemperature to which they reacted was 35-45° C, the indication was that a tempera-ture difference of about 15° C. was required to induce a positive response. It wasexpressed only as an accelerated rate of locomotion; on no occasion did the forelegsquest as in the shadow reflex. So the forelegs do not act as thermo-receptors likethe forelegs of the tick or the antennae of some insects.

Reactions to a linear temperature gradient

In investigating the reactions of small animals to a linear temperature gradient,MacLeod (1935) used a copper strip cooled at one end and warmed at the other.Wigglesworth (1941) used an apparatus consisting essentially of a zinc troughembedded deeply in a large trough containing sand. The zinc trough was cooledwith ice blocks at one end and warmed by means of a series of graded Bunsen flames.In the present work it was convenient to use a tubular linear temperature gradientapparatus.

A glass chamber, 27 cm. long and 1 cm. in diameter, was closed at one end witha rubber stopper, whilst a rubber stopper carrying a short length of narrow glasstubing was fitted into the other end. A thermometer was inserted into the chamberthrough the glass tube. The bulb of the thermometer could then be moved to anyrequired region of the chamber in order to record the air temperature.

The chamber was placed parallel to a window facing north. The closed end, fora distance of 4 cm., was embedded in ice, the opposite end rested upon a troughfilled with sand gently heated by a Bunsen flame. A temperature gradient extendedfrom 7 to 400 C, + 2° C, was maintained. A piece of filter-paper carrying aboutfifty mites was placed in the middle of the chamber. The mites had been previouslykept at 18° C. They quickly dispersed inside the chamber, but displayed avoidingreactions as they approached the lower and higher temperatures of the gradient. Inthe six experiments rough positional counts were made at intervals. The resultsindicated that there was not a well-defined preferred zone, but most of the miteswere aggregated between temperatures of 15 and 26° C. at any one time (see Table 1).

Table 1. Average distribution of fifty harvest mites when offered a linear temperaturegradient, graded from 7 to 400 C , ±2° C, inside a glass tube 27 cm. long and1 cm. in diameter.

Temp. (" C.;Average no. of mites

7-10

2

IO-II

411-15

616—20

26

20-25

11

25-3°1

30-400

Reactions to a concentric temperature gradient

The purpose of the tests was primarily to discover the nature of the movements ofthe mites away from the lower and higher temperatures of a gradient, because it wasnot easy to interpret these movements inside a glass tube.


A ground-glass plate, 30 cm. in diameter, was placed upon a peripheral circle ofice blocks. A blackened 25 W. bulb was placed below the centre of the under-surfaceof the glass plate. The ice reduced the temperature at the edge to about 70 C, whilsta temperature of about 40° C. was maintained at the centre. The temperatures weremeasured approximately by placing the bulb of a thermometer against the glass plate.

The mites were released at the centre of the experimental field. They moved fromthe centre towards the edge of the field. On approaching the cooler part they dis-played avoiding reactions. Similar reactions were observed in the central warmerregion. It was noticeable that as a mite moved into the cooler part the rate of locomo-tion decreased, but it was never immobilized or failed to turn round and walk in theopposite direction towards the warmer regions. On approaching the warmer partsthe mite displayed a detectable increase of pace before turning round and crawlingback down the gradient. The turning away from the cooler or warmer parts of thegradient was usually gradual and seldom abrupt. Moreover, the recorded paths ofthe mites clearly showed that as the mites moved around the circular field they didso within a well-defined zone owing to a series of avoiding reactions to the lower andhigher temperatures of the concentric gradient (see Fig. 17).

Fig. 17. Response of harvest mites to a concentric temperature gradient.

RESPONSE TO CHEMICAL STIMULATION

Henschel (1929) investigating the olfactory sense of Tyrolichus casei, the cheese mite,showed that it reacted positively to putrefying protein and one of its constituentsskatol, provided the concentration was not too high. The cheese mites reached baitssuch as meat juice and squashed caterpillars by very devious paths up the gradient ofintensity. When the tick Ixodes ricinus is stimulated chemically it starts waving itsfront legs (Totze, 1933). The front legs bear Haller's organs which are the chemo-receptors. Totze (1933) also showed that the tick was capable of picking out anoptimum zone in a gradient of butyric acid; Solomon (1937) found that varioussubstances were repugnant to Halotydeus destructor, the red-legged earth mite. It is

474 B. M. JONES

difficult to decide whether a mite has an olfactory sense, for although many workersassume that blood-sucking insects have a sense of smell and are attracted to the hostby the odour emanating from the body, difficulty arises in discriminating betweenan olfactory sense and a sense in the nature of taste.

The very excitable and purposeful movements of harvest mites crawling on thehairs or naked skin of a host suggested that odour was a possible stimulant. Theirpredilection for certain habitats on the host itself further suggested that odour mightexert a guiding influence. Host tissues and secretions as well as various volatilesubstances were therefore tested on the mites.

Reactions to volatile substances

Di-methyl-phthalate and di-butyl-phthalate were most successfully used duringthe military campaigns in the Far East for repelling Trombiculid mites. Certainother substances commonly used in the laboratory were found to be repugnant tothe mites, and these were used in tests designed to assess their sensory perception ofchemical stimulation.

Fifty mites were transferred to a glass tube 15 cm. long and 1 cm. in diameter.A small pad of cotton-wool was moistened with the test substance and placed insidethe tube near one end, being separated from the mites by a fine lawn screen sup-ported by a rubber ring. The open ends of the tube were plugged with cotton-wool(see Fig. 11). Light was directed upon the end containing the test substance so thatthe mites, being photopositive, were induced to move towards it.

It was found that the mites displayed avoiding reactions on approaching within0-5 cm. of phenol, methyl-phthalate, dilute ammonia, xylene and a 3% solution ofglacial acetic acid. Toluene appeared to have a stronger effect; the mites wererepelled at a distance of 1-5 cm. A mixture of 3 vol. of amyl acetate and 97 vol. ofwater was even more repellent, for the mites reacted to the vapour at a distance of5 cm. Mites approaching within this distance were narcotized by the vapour.Indifference was shown to glycerine and lactic acid.

The mites did not wave the forelegs in the manner described by Totze (1933) inhis observations on the reaction of ticks to chemical stimulation. Since the miteswere only able to detect the vapour at very short distance it must be assumed thattheir sensory perception of chemo-stimulation is not well developed and that highconcentrations of vapour are necessary to induce avoiding reactions. Although themite possesses sense organs of the chemo-receptor type (see p. 485) the entry of thenoxious vapour into the tracheal system which opens near the chelicerae may alsohave played a part in the perception of the volatile substances tested.

Reactions to host tissues and secretions

The host tissue or secretion to be tested as an attractant was placed in the centreof a filter-paper. The mites were released at various distances and from differentpositions around the test substance. The tests were conducted at a constant tempera-ture of 250 C.


Skin. Mites released at distances of 0-5 and 1 cm. from a piece of skin of a freshlykilled mouse were indifferent to its presence. Their tracks showed no relation to thepiece of skin (see Fig. 18 A).

Liver. The odour of a piece of fresh mouse liver proved equally negative (seeFig. 18 B).

Sebum and cerumen. The odour of hair emanates from the oil secreted by thesebaceous glands which are located around the hair roots. The odour of mouse,rabbit and human hair were tested on the mites but they showed no response (seeFig. 18 C). Complete indifference was also shown towards the odour of human andrabbit cerumen—the waxy secretion of the external meatus—which contains a highpercentage of sebum (see Fig. 18 E). Thus cerumen would not appear to play a partin attracting the mites into the ears, which are their characteristic habitats on rabbits.

Fig. 18. Tracks of harvest mites released at varying distances from host tissues and secretions at250 C. A, skin; B, liver; C, hair; D, perspiration; E, waxy secretion of external meatus.

Perspiration. A small pad of cotton-wool moistened with perspiration taken fromthe body of a dark-skinned person was placed on a filter-paper. Mites releasedat distances of 0-5 and 0-25 cm. showed complete indifference, but those which madecontact or near contact with the test substance displayed avoiding reactions (seeFig. 18D). Of 100 mites, placed five at a time, alongside the test substance, forty-six exhibited a negative response whilst the remainder were indifferent, althoughtwelve mites crawled on to the cotton pad for a brief period. The presence of decom-position products in the perspiration were probably responsible for repelling themites. Totze (1933) showed that a high concentration of butyric acid, which is oneof these products, repelled ticks. A pad of cotton-wool moistened in this substancewas repugnant to the mites.

The results were interesting because they help to explain why some people are nottroubled by the attentions of harvest mites. Dark-skinned people, who tend toperspire more than fair-skinned people, are less severely attacked by the mites.

476 B. M. JONES

Differences in the composition of the perspiration between one person and anotherare probably partly responsible for differences in the extent of an attack by the mites,because a few tests indicated that the perspiration of one person was found to bemore repugnant than that of another person. For example, only twelve out ofioo mites, placed in batches of five alongside a pad of cotton-wool moistened withperspiration taken from a fair-skinned person, displayed a well-defined avoidingreaction. However, the questions of differences in the degree of repulsion of perspira-tion of various hosts and of whether dark-skinned people are more immune thanfair-skinned people do not come within the scope of the present paper.

REACTIONS TO A YOUNG LIVE MOUSE

A young live mouse was chosen as the most suitable experimental host and placedin the centre of a filter-paper. Mites released at distances of 0-5, 1, 2 and 3 cm.showed complete indifference to its presence. Occasionally individual mites movedtowards the mouse and made devious paths under the legs and tail. Even then theyusually continued their tracks away from the mouse and towards the edge of thefilter-paper. Mites placed in close proximity with the body of the mouse alsoshowed indifference. But on climbing over the mouse and making contact with theskin the mite then displayed the striking behaviour so characteristic of a mite crawlingon the hand (see p. 481). It continued crawling at an accelerated pace upon the surfaceof the body for 30 min. or more (see Fig. 19). On reaching the distal parts of thelimbs or tail it as a rule showed no inclination to crawl off to the filter-paper butinstead clung tenaciously to the skin and retraced its tracks towards the body.

Fig. 19. Tracks of harvest mites released at varying distances from a young live mouse. The tracks,marked with a circle, led the mites to the body of the mouse, where they remained crawlingincessantly for a prolonged period.

It was evident from this behaviour that some stimulus other than touch wasresponsible for causing them to crawl incessantly over .the body of the mouse.However, in the odd case when a mite crawled off, it never turned round in anattempt to locate the mouse but instead made tracks towards the edge of the filter-paper. Occasionally a mite climbed back to the mouse after accidentally remaking


contact with an outstretched limb. It then crawled incessantly at an acceleratedrate over the body for a prolonged period. Mites previously kept on the body of themouse for 30 min. were placed at appropriate distances from the host, but they, too,showed complete indifference unless they touched it accidentally and climbed uponits body.

It was evident that the mites were incapable of detecting the presence of a younglive mouse even when they were almost touching it. But actual contact with the bodyof the mouse produced a response typified by an immediate increase of pace anda pronounced excitability. The body heat of the mouse, estimated at approximately35° C, probably accounted for the well-defined accelerated pace and pronouncedcharacter of the response. The behaviour closely resembled that displayed by themite crawling around the base of a glass tube warmed up to a temperature of35-40° c .

RESPONSE TO HUMIDITY

It is well known that compared with work on the reactions of animals to temperatureand light very few observations have been made on responses to humidity. In thenatural environment the mites take shelter from excess free water after heavy rain.On wet days any shelter is taken advantage of in the micro-habitat, although most ofthe mites cannot avoid being washed into the soil. The fact that mites will survivefor 4-5 days submerged in water suggests that heavy rains are not a serious handicapto them apart from restricting their movements. The importance of humidity asa limiting factor gave reason to believe that the mites were capable of perceivingvariations of the moisture content in the air or soil (Jones, 1950a).

Reactions to atmospheric moisture

Method. For investigating the reaction to a humidity gradient a glass tube 15 cm.long and 1 cm. in diameter was used. A moist pad of cotton-wool at one end andcalcium chloride at the other were separated from the introduced mites by plugs ofcotton-wool. Mites previously exposed either for 24 hr. at 100% R.H. and roomtemperature or for 6 hr. at 50% R.H. and 240 C. were placed in the humiditygradient in the tube.

Five tests were made with batches of fifty mites taken from saturated air. Theinitial phase was characterized by the introduced mites moving actively in all partsof the tube, but whereas they displayed well-marked avoiding reactions to the moistend they climbed upon the cotton pad at the dry end and even pressed themselvesbetween the cotton-wool and the glass tube. In one of the first tests the loose piecesof calcium chloride were not separated from the mite by a pad of cotton-wool. Inthis case the mites climbed upon the calcium chloride and even settled there.Avoidance of the moist end appeared to be responsible for most of the mites dis-playing an apparent preference for the dry end. Positional counts made at intervalsduring the first hour showed that most were recorded at the dry end (see Table 2).It was difficult to account for every mite because many had settled in the cotton-wool.After the first hour a few had settled on the moist pad, but most of them were still

IEB.27, 3 4 4 3 1

478 B. M. JONES

active and were wandering to and from the moist end. After 3 hr. there were stilla greater number of mites at the dry part, but the number of mites aggregated uponthe moist cotton-wool had also increased. After 6 hr. clusters of mites were formingon the moist cotton-wool and few mites were left upon the dry pad. After 24 hr.there was a complete reversal in distribution, practically all the mites having settledat the moist end (see Table 2), where many of them were by this time trapped incondensed moisture. Occasionally mites were trapped after penetrating the cottonpad at the dry end, and after a prolonged period of exposure they died throughdesiccation. Mites introduced into a tube closed at each end with a moist pad wereat first very active and avoided both ends, but after 3 hr. they settled and distributedthemselves more or less uniformly (see Table 2). The settled mites showed a ten-dency to cluster on or near the moist pads, but this could have been accounted forby their becoming trapped in condensed moisture.

Table 2. Distribution of fifty harvest mites, previously exposed for 24 hr. at I O O % R . H . ,

when offered alternate high and low humidities in a glass tube

Time

hr. min.

301 03 06 0

24 0

Alternative

100 % R.H.

Total

1112183846

Active

111 011

H2

humidities

2 5 %

Total

39383212

4

R.H.

Active

3S3429

91

Control

100%

Total

2 0

182 2——

R.H.

Active

1 0122 1——

100%

Total

303228——

R.H.

Active

141626——

Mites previously desiccated for 6 hr. at 50% R.H., 250 C. behaved in a humiditygradient differently from those taken from saturated air. They were very active afterrelease in the tube, but avoidance of the moist end was not so evident. After 1 hr.the mites, now much less active, were more or less uniformly distributed along thegradient. Only a few mites had climbed upon the pad at the dry end. After 3 hr.most of the mites had aggregated at the moist end and clusters had formed on themoist pad (see Table 3).

Observations on the behaviour of normal and desiccated mites in a humiditygradient suggested that the initial distribution of normal mites in the dry region is

Table 3. Distribution of fifty harvest mites, previously exposed for 6 hr. at 50% R.H.,240 C , when offered alternative high and low humidities in a glass tube (see text)

Time

hr. min.

153°

1 02 03 0

Alternative humidities

100% R.H.

Total

2627283638

Active

2625221812

25 % R.H.

Total

242322H12

Active

242318109


the result of an avoiding reaction to saturated air. Since desiccated mites did notshow this well-marked avoidance of the moist end, the expression of the avoidingreaction would appear to depend on the condition of humidity to which the miteshad previously been exposed.

The eventual aggregation, of normal mites after 24 hr. and desiccated mites after3 hr., at the moist end is due to a forced movement away from the dry region owingto water loss by evaporation through the cuticle. This depletion of the water balancestimulates the mites into activity, the random movements leading them into themoist region where movement is arrested whilst the water balance is restored.Hygrokinesis was more strongly displayed by the desiccated mites, which settledmore quickly in the moist region, than by those previously exposed to saturated air.This behaviour of the mites towards humidity closely resembles that of the tick(Lees, 1948) and the spider beetle, Ptinus tectus (Bentley, 1944).

Reaction to free water

A circle of filter-paper was wetted thoroughly at the periphery with a wet camel-hair brush. The water film 'crept' from the edge towards the centre. Mites werethen released at the centre of the filter-paper, and they displayed definite avoidingreactions to the surface film of the barrier ring of water (see Fig. 20).

Fig. 20. Response of harvest mites to free water. Barrier ring of water is representedby the area between the dotted line and the outer circle.

When the wetting operation was stopped and evaporation of the water from thefilter-paper reduced the amount of water in the outer circular area, the mites madedevious paths across the moist region towards the edge of the filter-paper.

Humidity and survival

Several workers, including Andre (1928) and Keay (1937), have emphasized theimportance of a high humidity for the successful rearing of the fed larva to thenymphal stage. Results of a previous investigation (Jones, 1950 a) suggested a relation-

31-2

480 B. M. JONES

ship between a high relative humidity and the distribution of the mites, but theresistance of the mites to dry conditions liable to be encountered in the naturalenvironment remained obscure. Mites were therefore exposed to various air humidi-ties to discover the most favourable one for survival.

Method. The mites were exposed to various humidities in a constant-temperatureroom at 240 C.

To produce the required humidity appropriate strengths of sulphuric acid wereintroduced into a series of 2 1. flasks. Six flasks were prepared with the followingrange of humidities: o, 10, 25,50,90, and 100%. Each flask was closed with a rubberstopper which carried a glass rod at the end of which was suspended a glass tube3^ cm. long and z\ cm. in diameter. The glass tube was closed above by a rubberstopper and below with a piece of lawn held tightly against the side of the tube bya rubber ring (see Fig. 13). The chamber successfully retained the mites. The flaskswere kept in the constant-temperature room for 2 days to allow the humidity in eachflask to attain equilibrium before the mites were introduced into the chamber.

Preliminary tests showed that hourly examinations were sufficient, because evenat 0% R.H. the mites remained alive for 6 hr. Twenty mites previously exposed tosaturated air were introduced into each chamber of the six flasks in six successivehumidity tests. When a close examination of the state of the mites was required, theglass tube was quickly removed from the flask and placed under a binocular micro-scope for as brief a period as possible, during which time the flask was stoppered.

The mites were at first very active at all humidities. After 3 hr. the movement ofmost of the mites in saturated air was arrested, but at the other humidities the miteswere still active. At the lower humidities, 10-50% R.H., they were continuallyactive until the depletion of water through evaporation proved fatal. Desiccationreduced the body size of the mites considerably at the end of 6 hr. at o % R.H. Thesustained activity was a response associated with the depletion of the water content.This activity was quite pronounced at 90% R.H., and the survival time wasappreciably less than that recorded for mites in saturated air. The average survivaltimes at different humidities are shown in Table 4. It was not difficult to recognizethe shrivelled appearance of a desiccated mite when assessing the mortality.

Table 4. Average time required to produce 100% mortality of harvest miteswhen exposed to varying degrees of relative humidity at 240 C.

Percentage relative humidityAverage time in hours required

to produce 100 % mortality

0

61 0

6251 0

SO

19

90

33 70

An atmosphere saturated with moisture was clearly the most favourable forsurvival. The appreciable resistance time of 6 hr. to o and 10% R.H. may have beendue to a use of the water phase of the tracheal system. The results of Davies (1928) inhis investigations on the effect of various humidities upon the survival of Collembolarevealed that tracheate species survived longer than atracheate ones.


REACTIONS TO TACTILE STIMULI

When a quiescent cluster of mites was touched lightly with a dry camel-hair brushthe mites immediately quested very actively and showed a remarkable ability forclimbing upon the hairs, from which it is difficult to remove them as they cling ontenaciously. The aroused mites will climb upon the wooden handle, and if given theopportunity will transfer themselves in a purposeful way to the fingers. If the miteis retained on the handle or brush part for a prolonged period, it stops morefrequently and gradually loses its initial excitability. The behaviour on the warmskin is quite different; with tenacity and at relatively great pace the mite climbsstraight from the hand to the arm whether it is held upward or downward. Thisresponse to contact with the warm skin is very striking. The photopositive responseis certainly obscured by it, since the mite will move from the exposed hand to thedarkness of the covered parts of the body—a curious reversal in behaviour towardslight in an animal whose unfed state has not changed.

When mites are placed upon the body of a rabbit or mouse they move withremarkable agility from one part of the body to another, and crawl at random forsome considerable time before they attach themselves to the skin. During thispreattachment phase they suffer a high mortality rate owing to the attentions of thehost. Observations of their movements upon the bodies of young mice suggestedthat a tactile sense was primarily responsible for the clusters upon typical sites ofa host. It is significant that inside the ears or around the anus or between the digitsare favoured habitats, and one would expect this because by their very nature as'niches' they are most likely to induce contact between the mites.

Mites often aggregate upon the corners of wooden planks or upon fallen branches.When the wooden plank or branch was lightly tapped the resulting vibrations inducedan immediate questing response.

CLUSTERING

The response to contact with each other's bodies is very noticeable in mites livingon the soil, and reference has been made in an earlier paper (Jones, 1950 a) to the partthis tactile response plays in the gregarious habit of forming clusters. But thephenomenon of clustering has been given further and separate attention in thispaper in its relation to the influence of light, humidity and gravity.

The influence of light

About fifty mites were introduced into a closed tube partly lined with moistfilter-paper. Occasionally a few mites were trapped in droplets of water. In thedark room the light of an overhead lamp induced the mites to climb upwards. Aftera prolonged period of 3 hr. or longer they formed a cluster at the top of the tube,behaviour resembling that of mites in a collecting jar exposed to daylight. A lampplaced at the side or below the glass tube induced the mites to form a cluster ata point nearest to the source of light. The upward-climbing.movement in such cases

482 B. M. JONES

was not apparent. The results suggested the upward-climbing action to be a responseto light intensity. A negative geotactic response, if it does exist, is most certainlyweaker than the photopositive response.

The influence of humidity

A wooden rod, supporting at half-way a circle of black filter-paper, 5 cm. indiameter, was glued at one end to the centre of the inside of a blackened Petri dish.Another equal-sized Petri dish, blackened except for a central circular area of about2 cm. in diameter, was inverted and placed upon the other so that the untreated areawas directly above the tip of the vertical wooden rod. They were held together witha strip of adhesive tape. Hence light entered the chamber through the circularwindow only, and it was conceivable that mites placed upon the suspended filter-paper within the chamber would respond to the light by climbing up the rod. Themites were introduced through a hole in the top Petri dish. The necessary strengthof acid was introduced into the chamber to give the required humidity. Thearrangement is shown in Fig. 12.

Fifty mites were used for each test, which lasted for at least 8 hr. and if necessaryuntil the following day. At 10 and 50% R.H. the mites were very active and climbedcontinuously up and down the rod. At 90% R.H. they remained longer at the tip,but no clusters were formed. At 95% R.H. and in saturated air mites aggregatedupon the tip and persistent clusters were formed at both humidities.

The influence of gravity

Several workers, including Hindle & Merriman (1912), MacLeod (1935) andLees (1948), have obtained varying results in testing the response of ticks to gravity.Lees (1948) concluded that ticks placed on a vertical rod displayed a form of negativegeotaxis, although occasionally they walked downwards for considerable distancesbefore turning, but if they had recently experienced arrival at the tip of the verticalrod, the turning upwards was enhanced. Since MacLeod (1935) and Lees (1948)found that ticks were photonegative (at variance with the conclusion of Totze (1933),who found them to be photopositive), an explanation for the upward-climbingaction of ticks had presumably to be sought in some form of gravity response.

In the case of harvest mites which are strongly photopositive, and sensitive todifferences of light intensity, it was not easy to test their responses to gravity. Onobserving mites in the field, one would be inclined to suggest that some form ofnegative geotaxis influences the upward-climbing action. But the response to lightcould equally explain this movement. Therefore gravity tests on the mites had to bearranged with the lighting either evenly distributed in the vertical plane or entirelyeliminated. Tests were arranged to determine the distribution of mites upona vertical rod in uniform lighting at 200 C, 95 % R.H.

The rod, 25 cm. long and 0-3 cm. in diameter, was a thin straight portion ofa raspberry cane, which in the natural environment is included among the favouredsites for climbing. The rod was marked at intervals of 1 cm. Glass rods were unsuit-able because they acquired a thin film of moisture which impeded the movement of


the mites. The rod was retained in position by two entomological pins fixed to anadjacent upright rod. Both rods were enclosed in a glass tube. The required humi-dity was obtained by lining the glass tube with filter-paper moistened with the appropriate salt solutionover part of its surface. This arrangement (see Fig. 21)resembled that of Lees (1948).

The humidity of the air inside the glass tube wasallowed to attain equilibrium before each test. Opaquescreens were arranged around the apparatus to shield .the glass tube from direct lighting, and the light in-tensity at the bottom and the top of the glass tube wasabout 300 m.c. The rod was withdrawn; the glass tubewas closed with a temporary stopper, and approximatelyfifty mites were quickly placed upon the rod at thehalf-way mark, before it was replaced.

Positional counts were then made at suitable inter-vals, but the total number varied at different intervals,either because some escaped notice, or others had foundtheir way across one of the pins to the supporting rod.Owing to this loss of mites, the total number at the endof a period of about 30 min. was invariably reduced.At the high humidity, one would expect that if some Fig. 21. Arrangement for test-form of negative geotaxis existed, the mites would climb in? the resP°nses of harvest

, . r « . 1 • i t mites to gravity at a high

to the tip of the rod, and on contacting each other humidity, a, testing rod; b,would aggregate as a cluster. The average results of pin ;c, supporting rod ;d, filter-eight tests are shown in Table S . p a p e r ; e< gIass tube"

Table 5. Average positional counts, expressed as percentage, of harvest mites distributedupon a vertical rod exposed to uniform lighting at 200 C , 95 % R.H.

_d

Section ofvertical rod

Top 1234

Bottom s

Duration from start of experiment

5 min.

3i-5i8-518-517-014-5

10 min.

26-516-0x8-517-022'O

15 min.

2O-Ox6-si6-s16-530-0

20 min.

25-017-012-512-533'°

30 min.

26-013-013-019-029-0

60 min.

n-o27-0

6-o18-038-0

The size and activity of the mites made it difficult to account for the total numberduring each count, but the figures, expressed as a percentage of each total count,indicate that when the mites are released upon the rod they are more inclined toclimb up rather than down the vertical rod during the first 5 min. After this periodthe counts suggest that the mites distribute themselves fairly evenly along the rod.It was noticed, too, that some were inclined to walk round and round the rod withina restricted range of a few centimetres. On no occasion did the mites form a per-

484 B. M. JONES

manent cluster at the tip of the rod, although some tests were carried out in darknessfor a period of 30 min. before the positions of the mites were examined.

The movements of individual mites were also followed. But the recorded trackswere so divergent in character that it was impossible to interpret them. The onlyfeature which had some consistency was an initial tendency on the part of the miteto climb upwards rather than downwards or round and round one particular part ofthe vertical rod. One may conclude from the results of the tests that there is nospecific indication of a negative geotaxis.

THE SENSE ORGANS

The external characters of systematic value for Trombicula autumnalis have beendescribed by Hirst (1915). In doing so he naturally gave only brief descriptions ofthe external form of the ocular areas, various setae and their positions on the body,the scutum, and the appendages. Systematists have attached much diagnosticimportance to the ciliation of the scutum, especially to the so-called pseudostigmaticsetae which, in a somewhat misleading way, they have been inclined to designatespecifically as 'sensilla', as if the considerable number of other setae present were oflittle significance as sensory end-organs.

According to the literature no study has been made of the structure and functionof the various receptor organs of T. autumnalis, nor as far as I know has anything ofthe kind been attempted for any species of Trombiculid mite. Unfortunately, theminute size of the harvest mite is not conducive to carrying out elimination experi-ments, designed on the lines of those by Wigglesworth (1941), to locate the senses.Attempts that were made did not produce any satisfactory results, excepting thoseof the blinding experiments. But the analysis of the sensory perceptions of the mitejustified a detailed examination of the sense organs, and since it had been possibleto measure in some degree the intensities of the responses of the mite to variousstimuli, it was desirable to try to identify the nature of the stimuli which the differentreceptor organs were capable of appreciating.

Sense organs in the legs

Plumose setae. These sensilla are confined mainly to the dorsal and lateral surfacesof the segments. Whether long or short they are slightly curved and taper to finepoints from a base which measures about 2 jii in diameter.

In section this type of receptor is shown to arise from a chamber-like structurewhich penetrates the cuticle. In some sections there appeared to be ridges on theinside of the chamber (see Fig. 22 D). The sections were stained with haematoxylin,but whereas a sensory nerve in the thick-walled hair was well defined, it was difficultto make out a sensory process in the chamber which would link it with the nerve cellsbelow the chamber itself. The nerve cells and their fibres associated with the sensillawere quite pronounced, but no structure resembling a trichogen or tormogen cellwas visible in the harvest mite. The structure and the wide distribution of the plumosehairs are typical features of sensilla with a tactile function.


Plain setae. There are fewer of these sensilla and they are located only on the legsand palps. A conspicuous one, about 80/x long, is present near the proximal end ofthe tarsus of the third leg. Their structure (see Fig. 22 E), apart from the lack ofbarbs, is similar to that of the plumose setae, which suggests that they too are tactilesensilla.

Fig. 22. A, dorsal view of right palp and first leg; B, detail of peg organ with round tip; C, detail oftactile sensillum and part of trifurcate claw; D, detail of plumose hair of body. E, detail of firstleg. a, plumose hair; b, plain hair; c, peg-organ with round tip; e, small rod;/, claw; g, sense cell;h, hypodermis; i, cuticle;j, nuclei of sense cells; k, sensory nerve; /, peg organ with pointed tip;m, tibia; n, patella; o, basi-femur; o', telo-femur; p, tarsus.

The peg organs. These organs are quite distinct in shape. There are two types,one with a round tip, the other with a pointed tip. A relatively stout slightly curvedpeg organ with a round tip is present on the tarsus of the first leg (see Fig. 22A).Those with pointed tips are more curved and slender.

In section the peg organ is shown to be thin walled (see Fig. 22 B), with a constric-tion at the base, so that a narrow aperture connects the cavity of the peg with that ofthe basal chamber which penetrates the cuticle. Below the chamber are sense cells

486 B. M. JONES

forming a nerve strand that can be traced to the chamber. The cavity of bothchamber and peg contained a clear fluid-like substance. A trichogen cell was notassociated with the peg organ.

The delicate thin walls, the spacious cavity filled with the vitreous substance andthe elongated sense cells closely resemble the peg organs regarded by Wigglesworth(1939) as chemo-receptors. Since its chemical sense is poorly developed, one wouldnot expect the mite to be well equipped with chemo-receptors. On the other hand,its perception of high or low temperatures is quite pronounced, and although thesensitivity to temperature differences is probably distributed over the body of themite, a possibility is that the pointed peg organs may be temperature receptors.

Minute sensory rods. These minute sensory organs are about 6/x long and 0-5 ju. indiameter (see Fig. 22E). They arise from a chamber comparable in size to that of thelarge peg organ. It was extremely difficult to be certain whether or not the end-organis thin walled. However, there is no sign of a sensory nerve enclosed by thick walls.They are principally confined to the first leg. There is one on each of the threedistal segments, and one is also present on the tarsus of the second leg. In someinsects such rods are grouped inside sunken epidermal pits and they have an olfactoryfunction.

Claws. The trifurcate claw, at the end of each tarsus, consists of a basal part andthree very curved distal branches, the middle one being the longest. The proximalpart of each branch is dilated (see Fig. 22 E). In section each branch is shown to bethick walled and supplied with a sensory nerve extending to the tip. Within the baseof the claw the sensory nerves to the three branches are separate, and each exhibitsa dilatation before joining the main nerve of the leg (see Fig. 22 C).

Sense organs on the palps

Most of the sensilla are confined to the papilliform tarsus. Two peg organs arepresent, one with a round tip and the other with a pointed one (see Fig. 22 A). Theconcentration of both tactile and chemo-receptor organs on the tarsus (see Fig. 23 C)is to be expected, since the palps play an important role in testing the skin before thedigits of the chelicerae are buried in it.

Sense organs elsewhere on the body

Plumose setae. Both the dorsal and ventral surfaces of the body are covered withthese sensilla. About thirty are arranged in transverse rows on the dorsal surface,excluding those on the scutum, and about the same number are similarly arrangedon the ventral surface.

Pseudostigmatic setae. The so-called pseudostigmatic setae usually arise froma mid-position of the scutum. In related genera these organs are club-shaped andclosely resemble the similarly termed organs, typical of members of the Oribatidae.

The club-shaped organs of orbatid mites by the nature of their structure werethought by Michael (1883) to be connected with hearing or smell, and he inclined tothe former.


The pseudostigmatic organs of T. autumnalis bear no resemblance to the club-shaped organs mentioned. They are, however, quite distinct from the rest of thesensilla so far described. The hair itself is about 70 m long, very slender, curved andequipped with barbs on the distal portion. The hair arises from a relatively largebasal process which fits into the space of the cuticular covering like the ball part ofa ball-and-socket joint. In section the basal process is seen to have a spaciouschamber about 8 /u. in diameter, which is enclosed by the convex dorsal surface anda pronounced concave ventral surface. The inside wall of the chamber is distinctlyridged (see Fig. 23 D). The significant feature is that this basal process, which in

Fig. 23. A, sagittal section of photoreceptor through anterior and posterior eyes; B, transversesection of anterior eye; C, horizontal section of right palp; D, transverse section of 'pseudo-stigmatic' organs, a, basal part of 'pseudostigmatic' organ; b, cells below lens of posterior eye;c, black pigment; d, lens; e, red pigmented oily substance; / , pigment cup; g, sense cells;h, hypodermis; i, cuticle; j , nuclei of sense cells; k, 'pseudostigmatic' hair; /, outer cortex ofsupra-oesophageal ganglion; m, ridge.

transverse section looks like an inverted cup, would appear to be capable of a rotatingmovement within the socket. Furthermore, this basal process projects into the bodycavity. Below each basal process is an aggregation of nerve cells with a thick strandextending to the ventral concave wall. These aggregations of nerve cells are quiteclose to the dorsally extended parts of the supra-oesophageal ganglia (see Fig. 23 D).

The end-organ is not thin walled, and thus it would seem that when it touchessome object the pressure exerted may cause a rotatory movement of the basal process.These stiff setae in their natural position extend well above the other dorsal setae ofthe body (see Fig. 14). One would therefore expect them to have a sensory percep-tion of touch and probably of vibrations in the air.

B. M. JONES

The ocular areas

The ocular areas lie on each side of the scutum near the postero-lateral border. Inthe live mite they appear as two conspicuously red oval patches, but each area iscomposed of two well-defined eyes, an anterior and a posterior one.

In sagittal section the two distinct eye structures are clearly visible. The cuticle ofthe anterior one is thickened into a distinct elliptical lens. Below the lens is a well-defined discoid mass of red-pigmented oily substance surrounded by a pronounceddark-coloured, almost black, pigmented cup. The posterior eye is not so welldeveloped. The lens-like cuticle is only about twice the thickness of the normalcuticular covering and half that of the lens of the anterior eye. In a transversesection of the anterior eye, the discoid mass of red-pigmented substance is continuouson both sides, with laterally placed nerve cells (see Fig. 23 B).

It is noticeable that the anterior eye will be exposed to rays of light from manydirections, but the posterior one is restricted in its range (see Fig. 23 A). In the caseof the better developed anterior eye one may assume that the thickened cuticularcovering, acting as a lens, brings the longer light rays to focus on the wall of thepigment cup; and that the shorter light rays after being reflected from the innersurface of the cup are focused on the photosensitive substance.

IDENTIFICATION OF THE RESPONSES

The responses of the hexapod mite to various stimuli have been identified, wherepossible, according to the classification of Fraenkel & Gunn (1940), because theadoption of some recognized scheme is essential if the expression of results of workon the sensory physiology of invertebrates is to be consistent.

The peculiar questing response of the harvest mite to a sudden decrease of lightintensity (shadow reflex), or to vibration, is, like that of the tick, a postural and nota locomotor activity. It cannot therefore be identified with either the random(kineses) or the directed locomotor reactions (taxes) which distinguish the twobasic groups of the classification.

Random movements

A random movement or kinesis involves the stimulation of an animal into loco-motor activity, but the body axis of the animal shows no orientation towards oraway from the stimulus.

Observations in the field showed that mites of a cluster, after questing in responseto shading or vibrations, invariably moved away from the site of the cluster. Thequesting response of the mite is a prelude to locomotion, and in this respect itclosely resembles the behaviour of the tick (Lees, 1948). The forward movementresulting from the stimulatory effect of shading or vibrations is an example oforthokinesis. The activity of desiccated mites in a linear humidity gradient is typicalof orthokinesis, the undirected movements eventually leading the mites into themoist region where movement is arrested. The response may also be defined aspositive hygrokinesis.


The extent of stimulation by contact with each others bodies leads to low activity,a quiescent state being typical of mites in a cluster. When only the tarsi of mitestouch a surface it leads to a high state of activity. On the surface of a collecting jar,or that of the human skin, the intensity of contact stimulation is low because only thetarsi make contact with the surface. The result is a high state of locomotor activityby the mite. This type of tactile response is an example of a low thigmokinesis,because a low intensity of stimulation by contact leads to a high state of activity. Thehigh extent of contact stimulation experienced by the mite pressed between the skinand tight clothing of a human host reduces its locomotor activity. The mite is theninduced to plunge the digits of the chelicerae into the skin, thus areas of the bodywhere clothing is tight are typical sites of attack.

The activity of the mite moving at random over the skin of a young live mousecould be interpreted as an example of klinokinesis in response to the warmth of thebody. On one occasion a mite traced a much convoluted path in response to twoopposing beams of different light intensity (see Fig. 3). But a response of this typewas not observed to chemical stimuli or temperature.

Direct movements

Unlike the random movement of the kinesis, a direct reaction or taxis involves thelong axis of the body of the animal being orientated in line with a single source ofstimulation. A positive or negative response depends on the mite moving towardsor away from the stimulus.

The photopositive mite will move along the bisector of two equal intersectinglights before curving towards one of the sources, and when one eyespot is blinded itwill make circus movements. Such responses are regarded as denoting phototropo-taxis. When the mite is placed in a weak light the tracks at a distance from the sourcestimulate klinotactic behaviour (see Fig. iA). As the mite approaches the sourcethe tracks straighten and tropotaxis is well defined. The reaction is more pronouncedin strong light and is also evident in the orientation of the mite towards a laterallypresented beam of light. The response to reflected light is an example of skototaxis.

The mite is unable to locate a warmed glass tube or the body of a young live mouse.The avoiding responses to volatile substances, perspiration, high humidity, thedarkened portion of a linear light-intensity gradient, and low and high temperaturesoutside a preferred zone, are all examples of negative responses to various stimuli.

THE SIGNIFICANCE OF THE RESPONSES IN THE NATURALENVIRONMENT WITH RESPECT TO ACQUIRING A HOST

The analysis of the sensory perceptions of the mite suggests that its behaviour in thenatural environment is a series of simple responses to stimuli which form the com-plex pattern of the micro-habitat. It is clear, however, that the intensity of responseto different stimuli varies, the response to mammal skin, for example, dominatingthe positive response to light when the combined stimuli are offered.

One would expect the behaviour of a potential parasite living freely on the soil tobe designed for acquiring a host, and hence that the stimuli encountered by the unfed

49° B. M. JONES

mite in the natural environment would be of value in this respect. When dry condi-tions prevail the locomotor activity of the mites over the surface of the soil or low-lying vegetation is almost exploratory in character, but there is nothing to suggestthat this behaviour is directed towards seeking a host as suggested by some observers.In describing the movements as representing a trial and error method it would bereasonable to infer the activity of the mite to be a series of avoiding reactions, themite being led into more favourable situations by moving away from those whichinduce a negative response. This behaviour would explain the restriction of mites tothe moister pockets of the macro-habitat.

The mites are unable to withstand prolonged exposure to low humidities and arethus confined to moist habitats. This is due to the high permeability of the cuticle towater, since mites soon shrivel up if kept in a dry Petri dish exposed to the heat ofstrong sunlight. Depletion of the normal water content stimulates the mite toactivity in the natural environment, locomotor activity being maintained until themite enters by chance a moister part where the rate of movement is slowed down.When the macro-climate is dry desiccation certainly prevents the mites from climbingupwards very far from the soil. Should they climb a twig or dead leaf they encounterdry air which induces a high locomotor activity, and this will lead them back to themoist soil to replenish any loss of water by evaporation. It is significant that mitesare difficult to locate when the soil is dry.

We may therefore deduce that the chances of a mite encountering a host areconsiderably lessened in dry conditions. Butwhen climatic conditions are favourable,for example, overnight rain followed by spells of warm sunshine, the combinedstimuli of warmth, fluctuating light and high humidity induce the mites in themicro-habitat to form clusters. The warmth stimulates the mites into activity, thesunlight induces the upward-climbing response, and the high humidity of the airallows the mites to reach the uppermost parts of the soil, dead leaves, twigs and low-lying vegetation without losing water by evaporation. The mites are drawn up fromthe soil, but a single mite will seldom remain alone upon the uppermost point of itsimmediate surroundings. It will climb down and then up to the tip of an adjacentrise in the soil, a twig or dead leaf. Mites of a concentrated pocket will, however,usually encounter each other at high points of the micro-habitat. On touching eachother they become still and form a quiescent cluster. Individual mites join up at thefringe until the cluster in parts may be formed of two or three layers of mites. It wasnoticeable that warm spells of sunshine, provided the humidity of the air was favour-able, invariably induced clustering. The aggregated mites usually rested upon a lumpof soil, a twig, or a leaf in a direct line with the sun, which often meant that clusterswere not always formed upon the uppermost points available on the material sup-porting them. Such behaviour is readily demonstrated in the laboratory by mitestransferred to a piece of crumpled filter-paper illuminated from the side by a lamp.The mites aggregate in positions, nearest to the source of light, irrespective of whetherthey are the highest points available. In the natural environment the mites showeda predilection for resting upon the sharp edges of the supporting material.

The formation of clusters at some distance above the soil would give the mite the


best opportunity of encountering a host. It is significant that clusters, once formed,will persist after dusk. The gradual decrease of light intensity has no influence uponthe mites, since the high intensity of stimulation by contact with each other keepsthe cluster intact. This persistence after dusk synchronizes with the movements ofrabbits, field mice, bank voles and other common hosts mainly nocturnal in theirhabits. During daytime, birds, domestic animals and some wild animals are thepotential hosts. But the mites on attaining a favourable position must still depend onthe chance approach of a suitable host.

The response either to shading or to vibrations is clearly one of preparedness.A potential host casts a shadow across a cluster of mites, and a quiescent clusterwould be transformed into a mass of questing appendages. Shading will also stopa mite in its tracks; the body rests on the substratum and all the legs are directedupwards. The questing response of mites to vibrations may have great value at nightwhen leaves and twigs, infested with clusters, are disturbed by nocturnal hosts.

Certain sensory perceptions which might be of value in locating a warm-bloodedhost are not possessed by the mite. It is incapable of orientating towards a host bydetecting the warmth or odour emanating from the body; but on touching a movingobject, inanimate or otherwise, the mite displays a remarkable ability for clinging toit. If the mite touches the skin it shows a well-defined response to body heat asshown by tests with a live mouse. The intensity of the response to the skin of thehost is so strong that the natural positive response of the mite to light is submerged,the avoiding reaction to darkness entirely disappearing. Hence the readiness of themite to move unhesitatingly to the covered parts of the human body. The role of theolfactory sense probably becomes more evident on close contact with the skin, andthe negative response to perspiration, varying in intensity according to the host,partly explains the apparent preference by the mite for one person more thananother.

On the host itself the mites maintain a high state of activity for a varying period oftime. During this pre-attachment phase there is a heavy mortality rate owing toscratching by the host. Those mites survive which eventually reach and attachthemselves to parts of the body where they escape the attentions of the host.Keay (1937) gives a list of habitats on both mammal and bird hosts: inside the ears,between the digits, and around the anus being typical sites. The ankles, groin, waist,axillae and neck are favoured regions of attack on the human body. These are areaswhere the clothing is usually more tightly pressed against the skin, and the mite insuch areas, by moving in a restricted space, becomes still in response to a highintensity of stimulation by contact. The palps are applied to the surface of the skin,the body is tilted and the cheliceral digits are buried into the horny layer (Jones,19506).

The mite feeds for about 3 days before detaching itself. The engorged mite climbsover the surface of the body before eventually dropping to the soil. Its furtherexistence depends entirely upon the moisture content of the soil. In the laboratoryengorged mites transferred to the surface of soil penetrate downwards; thus it isconceivable that under favourable conditions little or no lateral movement on the

492 B. M. JONES

surface of the soil precedes this apparent photonegative and photopositive geotacticbehaviour. The return to the cool soil from the warmth of the body is evidence ofa complete reversal of behaviour when the physiological state of the mite has changed.This penetration down into the soil is most probably a response to humidity, becauserecently fed mites are photopositive (see Fig. 24).

Fig. 24. Paths of fully engorged harvest mites, which had been attached to a host 6 hr. previouslyin a horizontal beam of light of 2600 m.c. (->) at the source.

Development to the nymphal stage will only take place under very moist condi-tions, and since the transition stage exists, for a period of about a month undernatural conditions, as a form incapable of locomotion, it is understandable that deeppenetration of the soil is essential to counteract desiccation. It is one of the strikingfeatures of the life cycle of T. autumnalis that the ectoparasitic larva is followed bya nymph and adult which lead a permanent hypogeal existence below the surface ofthe soil.

SUMMARY

Responses to stimuli

Light. In a strong beam of light the harvest mite will move directly towards thesource, whereas in a weak light the tracks are at first inclined to be wavy, but as theyapproach the source the tracks straighten. The mite moves along the bisector of twointersecting lights of equal intensity, and when blinded on one side makes circusmovements.

When offered a linear gradient of light intensity the mite avoids the darkenedportion and moves towards the lightest part of the field. Its movement towardssunlight is a true response to light and not to heat. A sudden decrease of lightintensity produces a questing response.

Temperature. The sensory perception of heat is poorly developed. The mite isincapable of locating a warm tube or the body of a young live mouse. On touchinga heated object it displays a well-defined response to a temperature difference ofabout 15° C. In a linear or concentric temperature gradient it displays avoidingreactions to low and high temperatures and appears to prefer a range extending from15 to 260 C.

Chemical stimulation. The mite is repelled at a distance of 0-5 cm. from phenol,methyl phthalate, dilute ammonia, xylene and a 3 % solution of glacial acetic acid.


Toluene was repugnant at 1 *5 cm., whilst a mixture of amyl acetate and water repelledthe mite at 5 cm. Complete indifference was shown to the odour of skin, liver, sebumand cerumen, but perspiration induced an avoiding reaction.

Humidity. Depletion of the water content influences the response of the mite tohumidity. A desiccated mite is active in dry air and inactive in moist air, buta normal individual will settle in either moist or dry air, while avoiding saturated air.The mite requires high humidities for prolonged survival, but avoids free water.

Touch. Unfed mites are very sensitive to touch. The extent of stimulation bycontact with each other's bodies, which is regarded as high, immobilizes them, and itis primarily responsible for the quiescent state of a cluster of mites. When the stimu-lation is low, for example, when only the tarsi are in contact with a surface, themite responds by displaying a high state of activity. A mite lightly touched willimmediately quest, a response induced equally by vibrations of the substratum.

Clustering

The gregarious habit of the mites is primarily a response to the touch of eachother's bodies. When the humidity is within the range 95-100% R.H. light willinduce the mites to climb up a rod and form a cluster at the tip. Whether or notnegative geotaxis also plays a part, it is difficult to say, because the evidence suggestedthat the mite is independent of gravity.

Sense organs

There are three types of sensilla: (1) tactile sensilla, both plumose and plain;(2) peg organs; (3) minute sensory rods, principally confined to the first leg. Anelliptical lens, a discoid mass of red-pigmented oily substance, and a pronounceddark pigmented cup are conspicuous features of the better developed anterior eye ofeach ocular area.

Identification of the responses

Where possible the responses of the mite to various kinds of stimuli have beenidentified according to a recognized scheme of classification.

Behaviour in the natural environment

The responses to stimuli which the mite will encounter in the natural environ-ment, and their value with respect to acquiring a host, are discussed.

I wish to thank Prof. Ritchie for his helpful criticism and interest, and toacknowledge the friendly discussions I had with the late Dr Gross on this work.Miss Catherine Hay kindly helped to prepare some of my sectioned material.

j E B . 2 7 , 3 4 4 32

494 B. M. JONES

REFERENCES

ANDRE, M. (1928). C.R. Acad. ScL, Paris, 19, 842.BENTLEY, E. W. (1944). J. Exp. Biol. 20, 152.COCKINGS, K. L. (1948). Bull. Ent. Res. 39, 281.DAVIES, W. M. (1928). J. Exp. Biol. 6, 79.FRAENKEL, G. A. & GUNN, D. L. (1940). The Orientation of Animals. Oxford: Clarendon Press.HENSCHEL, J. (1929). Z. vergl. Physiol. 9, 802.HINDLE, E. & MERRIMAN, G. (1912). Parasitology, 5, 203.HIRST, S. (1915). J. Econ. Biol. 10, 73.JONES, B. M. (1950a). Parasitology, 40, 1.JONES, B. M. (19506). Parasitology, 40, 247.KEAY, G. (1937). J. Anim. Ecol. 6, 23.KRIJGSMAN, B. J. (1937). Arch, neerl. Zool. 11, 401.LEES, A. D. (1948). J. Exp. Biol. 25, 145.MACLEOD, J. (1935). Parasitology, 24, 123.MICHAEL, A. D. (1883). British Oribatidae, 1. London: Ray Society.OEHRING, W. (1934). Zool. Jb. S3, 342.RIVNAY, E. (1932). Parasitology, 24, 121.SOLOMON, M. E. (1937)- J- Anim. Ecol. 6, 340.TOTZE, R. (1933). Z. vergl. Physiol. 19, n o .WIGGLESWORTH, V. B. (1939). The Principles of Insect Physiology. London: Methuen.WIGGLESWORTH, V. B. (1941). Parasitology, 33, 67.WIGGLESWORTH, V. B. & GILLETT, J. D. (1934). J. Exp. Biol. 11, 120.

6i ] the sensory physiology of the harvest ...sensory physiology of the harvest mite trombicula...

Documents